The International Agency for Research on Cancer has judged that the evidence for the carcinogenicity of silica in animals is sufficient, and for humans it is limited. With the use of the Global 86 computer model, animal and human cancer potency data were extrapolated and the findings compared. The animal findings were based on inhalation rat studies. The cancer slope factors ranged from 2.3×10-5 to 6.0×10-3 for (1 µg·m-3)-1 among three experimental studies. The epidemiologic findings were from gold workers exposed to quartz and diatomaceous earth workers exposed to cristobalite. The cancer slope factors ranged from 6.8×10-7 to 1.85×10-5 for lifetime exposure to 1(µg·m-3)-1 of silica dust. Because of the many uncertainties involved in extrapolating to humans from animal data, more rational risk assessments are achieved when data from silica-exposed workers are used than when laboratory findings are relied on.